MVD has a very variable progression, and there is a question still unsolved: when must we start the treatment? For it's now well demonstrated that if ACE inhibitors improve clinical status and prolong longevity in dogs with symptomatic MVD, unfortunately they have no preventive effect in asymptomatic stages ; 3 successive publications demonstrated it now unquestionably: Kvart, SVEP trial 2002; Atkins, VETPROOF, 2007; Pouchelon, JVIM, 2008. Therefore, we evidently need a "key" to predict the risk and time to onset of CHF, just to permit a precocious intervention and then to really improve outcome. Many attempts were published, not really convincing, except one of the latest (Reynolds, ACVIM, 2008) using Left Atrium and Aorta ratio ( LA/Ao) and serum NT-proBNP: 45% probability of CHF when LA/Ao = 1,8 & NT-proBNP = 1086 ; 100% probability of CHF when LA/ Ao = 2,8 & NT-proBNP = 2192. Two shortcomings in this statistical model: 1) It provides a probability, but not a threshold. 2) Serum NT-proBNP dosage is not yet a routine one, immediately accessible to all practitioners. A few years ago (ICVS, Paris, 2004), we underlined the interest of the Left Ventricular Internal Diameter in diastole

(LVd) calculated on a standard transventricular M-mode as an interesting measurement correlated with many other parameters: murmur, class of CHF, LA/Ao (Häggström, JVIM, 1995), prognosis and mortality prediction (Moonarmat, JSAP, 2010). But because raw measurement are not available between dogs, we were very interested with the M-mode echographic ratio indices published by Brown et al (JVIM, 2003), and particularly with the LVd/Ao ratio: in normal dog, LVd/Ao = 1,6 +/- 0,2 , that means LV dilation starts over 1,8 (and then NT-proBNP secretion too).

Because La/Ao and LVd/Ao separately failed to discriminate asymptomatic and symptomatic MVD (classes 1a-1b from classes 2-3a), despite a good linear correlation of these 2 indexes with CHF classes, our idea was to add the 2 indexes: we realised it on 397 dogs with MVD, and the results seem to be promising: Left Chambers on Aorta ratio (LC/Ao) = LA/Ao + LVd/Ao = in Class 1a: 3,12; Class1b: 4,01; Class 2: 4,59; Class 3a: 5,04 (see tables for details). It means that if you have an asymptomatic MVD with a LC/Ao > 4,5, you may immediately start treatment with ACE inhibitors, Pimobendan, or even Spironolactone more usefully than with a blind approach.